jDJddlZddlmZmZddlmZddlmZddlm Z ddl m Z m Z m Z mZmZmZddlmZddlZddlmZmZmZd d lmZd d lmZmZmZmZmZm Z m!Z!m"Z"m#Z#m$Z$m%Z%m&Z&m'Z'd d l(m)Z)eGd dZ*eGddZ+eGddZ,Gdde-eZ.Gdde-eZ/dede-defdZ0GddZ1GddZ2Gdd Z3Gd!d"Z4Gd#d$Z5y)%N) dataclassasdict)Enum)chain)unescape)ListDictIteratorIterablePatternOptional) ElementTree) HTTPErrorSessionResponse) ProxyConfig) VideoUnavailableYouTubeRequestFailedNoTranscriptFoundTranscriptsDisabledNotTranslatableTranslationLanguageNotAvailableFailedToCreateConsentCookieInvalidVideoId IpBlockedRequestBlocked AgeRestrictedVideoUnplayableYouTubeDataUnparsable) WATCH_URLc.eZdZUeed<eed< eed<y)FetchedTranscriptSnippettextstartdurationN)__name__ __module__ __qualname__str__annotations__floatd/root/aria/tools/markitdown-venv/lib/python3.12/site-packages/youtube_transcript_api/_transcripts.pyr#r#"s I LOr.r#ceZdZUdZeeed<eed<eed<eed<eed<de efdZ defd Z de fd Z deefd Zy ) FetchedTranscriptz Represents a fetched transcript. This object is iterable, which allows you to iterate over the transcript snippets. snippetsvideo_idlanguage language_code is_generatedreturnc,t|jSN)iterr2selfs r/__iter__zFetchedTranscript.__iter__>sDMM""r.c |j|Sr9)r2)r<indexs r/ __getitem__zFetchedTranscript.__getitem__As}}U##r.c,t|jSr9)lenr2r;s r/__len__zFetchedTranscript.__len__Ds4==!!r.c>|Dcgc] }t|c}Scc}wr9)r)r<snippets r/ to_raw_datazFetchedTranscript.to_raw_dataGs/34Gw444sN)r'r(r)__doc__rr#r+r*boolr r=r@intrCr rFr-r.r/r1r11sf +,,MM#(#;<#$$<$""5T$Z5r.r1c"eZdZUeed<eed<y)_TranslationLanguager4r5N)r'r(r)r*r+r-r.r/rKrKKs Mr.rKceZdZdZdZdZy)_PlayabilityStatusOKERRORLOGIN_REQUIREDN)r'r(r)rNrOrPr-r.r/rMrMQs B E%Nr.rMceZdZdZdZdZy)_PlayabilityFailedReasonu%Sign in to confirm you’re not a botzSign in to confirm your agezVideo unavailableN)r'r(r) BOT_DETECTEDAGE_RESTRICTEDVIDEO_UNAVAILABLEr-r.r/rRrRWs:L2N+r.rRresponser3r7cb |j|S#t$r}t||d}~wwxYwr9)raise_for_statusrr)rVr3errors r/_raise_http_errorsrZ]s54!!# 4"8U334s . ).czeZdZdededededededeefdZdd ed e fd Z d efd Z e d efd Z ded dfdZy) Transcript http_clientr3urlr4r5r6translation_languagesc||_||_||_||_||_||_||_|Dcic]}|j|jc}|_ycc}w)z You probably don't want to initialize this directly. Usually you'll access Transcript objects using a TranscriptList. N) _http_clientr3_urlr4r5r6r__translation_languages_dict) r<r]r3r^r4r5r6r_translation_languages r/__init__zTranscript.__init__fsm(     *(%:")>, $ ! . .0D0M0M M, (, sApreserve_formattingr7c.|jj|j}t|j t ||j j}t||j |j|j|jS)z Loads the actual transcript data. :param preserve_formatting: whether to keep select HTML text formatting )rf)r2r3r4r5r6) ragetrb_TranscriptParserparserZr3r$r1r4r5r6)r<rfrVr2s r/fetchzTranscript.fetchsz $$((3$9LMSS x 7 < < !]]]],,**   r.cxdj|j|j|jrdSdS)Nz7{language_code} ("{language}"){translation_description}z[TRANSLATABLE])r4r5translation_description)formatr4r5is_translatabler;s r/__str__zTranscript.__str__sNHOO]],,8<8L8L$4P  SUP  r.c2t|jdkDS)Nr)rBr_r;s r/rpzTranscript.is_translatables4--.22r.c $|jst|j||jvrt |jt |j |jdj|j||j||dgS)Nz{url}&tlang={language_code})r^r5T) rprr3rcrr\rarorb)r<r5s r/ translatezTranscript.translates##!$--0 0  @ @ @1$--@ @    MM ) 0 0II] 1   , ,] ;     r.NF)r'r(r)rr*rHrrKrer1rkrqpropertyrprtr-r.r/r\r\es           $$89 4  :K "  333 s | r.r\c eZdZdZdedeeefdeeefdeefdZ e de deded dfd Z d e efd Zd eed efd Zd eed efdZd eed efdZd eedeeeefd efdZd efdZdeed efdZy)TranscriptListz This object represents a list of transcripts. It can be iterated over to list all transcripts which are available for a given YouTube video. Also it provides functionality to search for a transcript in a given language. r3manually_created_transcriptsgenerated_transcriptsr_c<||_||_||_||_y)a The constructor is only for internal use. Use the static build method instead. :param video_id: the id of the video this TranscriptList is for :param manually_created_transcripts: dict mapping language codes to the manually created transcripts :param generated_transcripts: dict mapping language codes to the generated transcripts :param translation_languages: list of languages which can be used for translatable languages N)r3_manually_created_transcripts_generated_transcripts_translation_languages)r<r3ryrzr_s r/rezTranscriptList.__init__s#! -I*&;#&;#r.r] captions_jsonr7c r|jdgDcgc]}t|dd|d}}i}i}|dD]f}|jddd k(r|}n|}t|||d |d d|d|jddd k(|jd d r|ng||d<ht||||Scc}w)a] Factory method for TranscriptList. :param http_client: http client which is used to make the transcript retrieving http calls :param video_id: the id of the video this TranscriptList is for :param captions_json: the JSON parsed from the YouTube pages static HTML :return: the created TranscriptList translationLanguages languageName simpleText languageCoder4r5 captionTrackskindrmasrbaseUrlnameisTranslatableF)rhrKr\rx) r]r3rrdr_ryrzcaptiontranscript_dicts r/buildzTranscriptList.builds ")6(9(9:PRT(U ! % !-n=lK2>B ! ! (*$ "$_5 G{{62&%/"7">7A " -' FB'50)05Eu)M%SU8OGN3 4     ( ! !   7! sB4czt|jj|jjSr9)rr|valuesr}r;s r/r=zTranscriptList.__iter__s3  . . 5 5 7  ' ' . . 0  r.language_codescR|j||j|jgS)a> Finds a transcript for a given language code. Manually created transcripts are returned first and only if none are found, generated transcripts are used. If you only want generated transcripts use `find_manually_created_transcript` instead. :param language_codes: A list of language codes in a descending priority. For example, if this is set to ['de', 'en'] it will first try to fetch the german transcript (de) and then fetch the english transcript (en) if it fails to do so. :return: the found Transcript )_find_transcriptr|r}r<rs r/find_transcriptzTranscriptList.find_transcripts.$$   / /1L1L M  r.c<|j||jgS)a Finds an automatically generated transcript for a given language code. :param language_codes: A list of language codes in a descending priority. For example, if this is set to ['de', 'en'] it will first try to fetch the german transcript (de) and then fetch the english transcript (en) if it fails to do so. :return: the found Transcript )rr}rs r/find_generated_transcriptz(TranscriptList.find_generated_transcripts $$^d6Q6Q5RSSr.c<|j||jgS)a| Finds a manually created transcript for a given language code. :param language_codes: A list of language codes in a descending priority. For example, if this is set to ['de', 'en'] it will first try to fetch the german transcript (de) and then fetch the english transcript (en) if it fails to do so. :return: the found Transcript )rr|rs r/ find_manually_created_transcriptz/TranscriptList.find_manually_created_transcripts%$$ T??@  r.transcript_dictscf|D]}|D]}||vs||ccSt|j||r9)rr3)r<rrr5rs r/rzTranscriptList._find_transcript(sJ , :M#3 : O3*=99 : :   ~tDDr.c 8dj|j|jd|jj D|jd|j j D|jd|j DS)NaFor this video ({video_id}) transcripts are available in the following languages: (MANUALLY CREATED) {available_manually_created_transcript_languages} (GENERATED) {available_generated_transcripts} (TRANSLATION LANGUAGES) {available_translation_languages}c32K|]}t|ywr9r*.0 transcripts r/ z)TranscriptList.__str__..?s[J[c32K|]}t|ywr9rrs r/rz)TranscriptList.__str__..CsK$.JKrc3jK|]+}dj|j|j-yw)z{language_code} ("{language}")rN)ror4r5)rrds r/rz)TranscriptList.__str__..Fs>K ) 1771::"6"D"D8Ks13)r3/available_manually_created_transcript_languagesavailable_generated_transcriptsavailable_translation_languages)ror3_get_language_descriptionr|rr}r~r;s r/rqzTranscriptList.__str__4s 0 &]]<@djd|D}|r|SdS)N c3@K|]}dj|yw)z - {transcript})rN)rors r/rz;TranscriptList._get_language_description..Ps&   $ $ $ ; sNone)join)r<r descriptions r/rz(TranscriptList._get_language_descriptionOs,ii 0  *{5v5r.N)r'r(r)rGr*r r\rrKre staticmethodrrr r=r rrrrrqrr-r.r/rxrxs4 <<'+3 ?&;< $CO4 < $$89 <(+ + (++ <@+ + + Z (:.  hsm  T T* T  &sm     E   EtCO45 E  E  66HSM6c6r.rxceZdZdedeefdZdedefdZ ddede de fdZ d edede fd Z d e dedd fd Zd ededd fdZdedefdZdedefdZy )TranscriptListFetcherr] proxy_configc ||_||_yr9)ra _proxy_config)r<r]rs r/rezTranscriptListFetcher.__init__Xs')r.r3r7cbtj|j||j|Sr9)rxrra_fetch_captions_json)r<r3s r/rkzTranscriptListFetcher.fetch\s/##      % %h /  r. try_numberc, |j|j||S#t$rg}|jdn|jj}|dz|kr|j ||dzcYd}~S|j |jd}~wwxYw)Nrr)r)_extract_captions_json_fetch_video_htmlrrretries_when_blockedrwith_proxy_config)r<r3r exceptionretriess r/rz*TranscriptListFetcher._fetch_captions_jsoncs B..&&x0(  B%%-''<<  A~'00jSTn0UU--d.@.@A A Bs! # BAB-B3BBhtmlc,td} |j||}|j |j d||j dij d}|d|vr t ||S#t$r}d|vr t||d}~wwxYw)NytInitialPlayerResponsezclass="g-recaptcha"playabilityStatuscaptionsplayerCaptionsTracklistRendererr) _JsVarParserrjr r_assert_playabilityrhr)r<rr3 var_parser video_dataers r/rz,TranscriptListFetcher._extract_captions_jsonrs!";<  #))$9J   0C!DhO"z26:: -   O=$H%h/ /% $,))G  sA44 B=BBplayability_status_dataNc .|jd}|tjjk7ra|]|jd}|tjjk(rP|t j jk(r t||t jjk(r t||tjjk(rU|t jjk(r8|jds|jdr t|t||jdijdijdijdg}t|||Dcgc]}|jd d c}yycc}w) Nstatusreasonzhttp://zhttps:// errorScreenplayerErrorMessageRenderer subreasonrunsr$rm)rhrMrNvaluerPrRrSrrTrrOrU startswithrrr)r<rr3playability_statusr subreasonsruns r/rz)TranscriptListFetcher._assert_playabilitysc488B "4"7"7"="= =".,00:F!%7%F%F%L%LL5BBHHH(225DDJJJ'11"&8&>&>&D&DD6HHNNN&&y1X5H5H5T(22&x00'++M2>126["%VR  "&*"M337762#6"M +/ >.#Ns0F ctjd|}| t||jjj dd|j dzdy)Nzname="v" value="(.*?)"CONSENTzYES+rz .youtube.com)domain)researchrracookiessetgroup)r<rr3matchs r/_create_consent_cookiez,TranscriptListFetcher._create_consent_cookiesV 2D9 =-h7 7 !!%% v A.~ & r.c|j|}d|vr2|j|||j|}d|vr t||S)Nz&action="https://consent.youtube.com/s") _fetch_htmlrr)r<r3rs r/rz'TranscriptListFetcher._fetch_video_htmlsR) 3t ;  ' 'h 7##H-D74?1(;; r.c|jjtj|}t t ||j S)N)r3)rarhr!rorrZr$)r<r3rVs r/rz!TranscriptListFetcher._fetch_htmls<$$(()9)98)LM*8X>CCDDr.)r)r'r(r)rr rrer*rxrkrIr rrrrrrr-r.r/rrWs*G*8K;P* c n  BS Bc B$ B3#$(43SW: 3 # $ ##ECECEr.rcNeZdZgdZd defdZdedeefdZdede e fdZ y) ri) strongembimarksmalldelinssubsuprfc0|j||_yr9)_get_html_regex _html_regex)r<rfs r/rez_TranscriptParser.__init__s//0CDr.r7c|rIdj|j}d|zdz}tj|tj}|Stjdtj}|S)N|z <\/?(?!\/?(z )\b).*?\b>z<[^>]*>)r_FORMATTING_TAGSrcompile IGNORECASE)r<rf formats_regex html_regexs r/rz!_TranscriptParser._get_html_regexs` HHT%:%:;M*]:]JMM2==AJJ >Jr.raw_datac Xtj|Dcgc]}|jytt j |j dt|jt|jdt|jjddc}Scc}w)Nrmr%durz0.0)r$r%r&) r fromstringr$r#rrrrr,attribrh)r<r xml_elements r/rjz_TranscriptParser.parses +55h?  + %VVD,,b(;;K;K2LMK..w78{1155eUCD    sB B'Nru) r'r(r)rrHrer r*rrr#rjr-r.r/ririsH EDE4GCL  c  d+C&D  r.ric`eZdZdefdZdededefdZdededeefdZdeededefd Z y ) rvar_namec||_yr9) _var_name)r<rs r/rez_JsVarParser.__init__s !r.raw_htmlr3r7ct|j||}|j||}tj|Sr9)_create_var_char_iterator_find_var_substringjsonloads)r<rr3 char_iterator var_strings r/rjz_JsVarParser.parses566xJ --mXF zz*%%r.c|jd|j}t|dkr t|t |d}t |dk7r t |dk7r|S)Nzvar r{)splitrrBr r:next)r<rr3 splitted_htmlr s r/rz&_JsVarParser._create_var_char_iteratorsh dnn-='>? }  "'1 1]1-. =!S( =!S(r.r cd}d}d}dg}|D]X}|j||rd}n(|dk(rd}n |dk(r| }n|s|dk(r|dz }n |dk(r|dz}|dk(sGd j|cSt|) NFrr\T"}rrm)appendrr )r<r r3escaped in_quotesdepthcharschars r/rz _JsVarParser._find_var_substrings ! &D LL  )M 3;QJES[QJEzwwu~% &"$H--r.N) r'r(r)r*rer rjr rrr-r.r/rrsf""&c&S&T& #RU.#.#.RU.r.r)6r  dataclassesrrenumr itertoolsrrrtypingrr r r r r defusedxmlrrrequestsrrrproxiesr_errorsrrrrrrrrrrrrr _settingsr!r#r1rKr*rMrRrZr\rxrrirr-r.r/r%s )DD" 11 !      55 52   &d& ,sD, 44S4X4H H Vd6d6N_E_ED# # L).).r.